Prediksi Tingkat Prokrastinasi Mahasiswa Berdasarkan Pola Penggunaan Media Sosial Menggunakan Algoritma Categorical Naïve Bayes

Abstract

Academic procrastination is the tendency to delay completing academic tasks and remains a common issue among university students. This study aims to develop a prediction model for students' academic procrastination levels based on social media usage patterns using the Categorical Naïve Bayes algorithm, evaluate the model performance using accuracy, precision, recall, and F1-score, and identify the variables that contribute the most to the prediction results. This research employed the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework. Data were collected through questionnaires distributed to 300 active students of the Informatics Engineering Study Program at Universitas Muhammadiyah Sukabumi. The predictor variables included academic year, gender, duration of social media usage, frequency of social media access, dominant time of use, the most frequently used social media platform, and the habit of checking social media notifications while studying. The results showed that the Categorical Naïve Bayes model achieved an accuracy of 35%. Evaluation using precision, recall, and F1-score indicated that the model still had difficulty distinguishing among the academic procrastination categories. Furthermore, Permutation Importance analysis revealed that notification-checking habits, time of social media usage, and social media platform were the variables contributing the most to the prediction process. These findings indicate that social media usage patterns can be utilized as predictor variables for developing an academic procrastination prediction model.